Exploring the anion chemical space of Ln2OF2−xClxH2 (Ln = Y, La, Gd): a model of electroelastic material with high mechanical sensitivity and energy harvesting

<div>Recent years have witnessed utilization of modern machine learning approaches for predicting properties of material using available datasets. However, to identify potential candidates for material discovery, one has to systematically scan through a large chemical space and subsequently calculate the properties of all such samples. On the other hand, generative methods are capable of efficiently sampling the chemical space and can generate molecules/materials with desired properties. In this study, we report a deep learning based inorganic material generator (DING) framework consisting of a generator module and a predictor module. The generator module is developed based upon conditional variational autoencoders (CVAE) and the predictor module consists of three deep neural networks trained for predicting enthalpy of formation, volume per atom and energy per atom chosen to demonstrate the proposed method. The predictor and generator modules have been developed using a one hot key representation of the material composition. A series of tests were done to examine the robustness of the predictor models, to demonstrate the continuity of the latent material space, and its ability to generate materials exhibiting target property values. The DING architecture proposed in this paper can be extended to other properties based on which the chemical space can be efficiently explored for interesting materials/molecules.</div><div><br></div>

Download Full-text

Deep Learning Enabled Inorganic Material Generator

10.26434/chemrxiv.12312260 ◽

2020 ◽

Author(s):

Yashaswi Pathak ◽

Karandeep Singh Juneja ◽

Girish Varma ◽

Masahiro Ehara ◽

U. Deva Priyakumar

Keyword(s):

Deep Learning ◽

Deep Neural Networks ◽

Inorganic Material ◽

Chemical Space ◽

Property Values ◽

Learning Approaches ◽

Predictor Models ◽

Generative Methods ◽

Material Space ◽

Modern Machine

<div>Recent years have witnessed utilization of modern machine learning approaches for predicting properties of material using available datasets. However, to identify potential candidates for material discovery, one has to systematically scan through a large chemical space and subsequently calculate the properties of all such samples. On the other hand, generative methods are capable of efficiently sampling the chemical space and can generate molecules/materials with desired properties. In this study, we report a deep learning based inorganic material generator (DING) framework consisting of a generator module and a predictor module. The generator module is developed based upon conditional variational autoencoders (CVAE) and the predictor module consists of three deep neural networks trained for predicting enthalpy of formation, volume per atom and energy per atom chosen to demonstrate the proposed method. The predictor and generator modules have been developed using a one hot key representation of the material composition. A series of tests were done to examine the robustness of the predictor models, to demonstrate the continuity of the latent material space, and its ability to generate materials exhibiting target property values. The DING architecture proposed in this paper can be extended to other properties based on which the chemical space can be efficiently explored for interesting materials/molecules.</div><div><br></div>

Download Full-text

Relationships between hierarchical structure and properties in macromolecular systems

Proceedings, annual meeting, Electron Microscopy Society of America ◽

10.1017/s0424820100126950 ◽

1987 ◽

Vol 45 ◽

pp. 440-443

Author(s):

E. Baer

Keyword(s):

Uniaxial Tension ◽

Hierarchical Structure ◽

Inorganic Material ◽

Hierarchical Structures ◽

Bone Collagen ◽

Structure And Properties ◽

Connective Tissues ◽

Scale Level ◽

Fibrous Protein ◽

Macromolecular Systems

The most advanced macromolecular materials are found in plants and animals, and certainly the connective tissues in mammals are amongst the most advanced macromolecular composites known to mankind. The efficient use of collagen, a fibrous protein, in the design of both soft and hard connective tissues is worthy of comment. Very crudely, in bone collagen serves as a highly efficient binder for the inorganic hydroxyappatite which stiffens the structure. The interactions between the organic fiber of collagen and the inorganic material seem to occur at the nano (scale) level of organization. Epitatic crystallization of the inorganic phase on the fibers has been reported to give a highly anisotropic, stress responsive, structure. Soft connective tissues also have sophisticated oriented hierarchical structures. The collagen fibers are “glued” together by a highly hydrated gel-like proteoglycan matrix. One of the simplest structures of this type is tendon which functions primarily in uniaxial tension as a reinforced elastomeric cable between muscle and bone.

Download Full-text

Analysis of magneto rheological elastomers for energy harvesting systems

International Journal of Applied Electromagnetics and Mechanics ◽

10.3233/jae-209350 ◽

2020 ◽

Vol 64 (1-4) ◽

pp. 439-446

Author(s):

Gildas Diguet ◽

Gael Sebald ◽

Masami Nakano ◽

Mickaël Lallart ◽

Jean-Yves Cavaillé

Keyword(s):

Magnetic Field ◽

Energy Harvesting ◽

Shear Strain ◽

Magnetic Induction ◽

Magnetic Particles ◽

Homogeneous Magnetic Field ◽

Electrical Signal ◽

Harvesting Systems ◽

Dc Bias ◽

Magneto Rheological

Magneto Rheological Elastomers (MREs) are composite materials based on an elastomer filled by magnetic particles. Anisotropic MRE can be easily manufactured by curing the material under homogeneous magnetic field which creates column of particles. The magnetic and elastic properties are actually coupled making these MREs suitable for energy conversion. From these remarkable properties, an energy harvesting device is considered through the application of a DC bias magnetic induction on two MREs as a metal piece is applying an AC shear strain on them. Such strain therefore changes the permeabilities of the elastomers, hence generating an AC magnetic induction which can be converted into AC electrical signal with the help of a coil. The device is simulated with a Finite Element Method software to examine the effect of the MRE parameters, the DC bias magnetic induction and applied shear strain (amplitude and frequency) on the resulting electrical signal.

Download Full-text

Simply supported FPEDs connected by springs for broadband energy harvesting

International Journal of Applied Electromagnetics and Mechanics ◽

10.3233/jae-209323 ◽

2020 ◽

Vol 64 (1-4) ◽

pp. 201-210

Author(s):

Yoshikazu Tanaka ◽

Satoru Odake ◽

Jun Miyake ◽

Hidemi Mutsuda ◽

Atanas A. Popov ◽

...

Keyword(s):

Energy Harvesting ◽

Evaluation Method ◽

Energy Harvester ◽

Vibrational Energy ◽

Functional Materials ◽

Vibration Energy ◽

Theoretical Evaluation ◽

Vibration Energy Harvester ◽

Polyvinylidene Difluoride ◽

Simply Supported

Energy harvesting methods that use functional materials have attracted interest because they can take advantage of an abundant but underutilized energy source. Most vibration energy harvester designs operate most effectively around their resonant frequency. However, in practice, the frequency band for ambient vibrational energy is typically broad. The development of technologies for broadband energy harvesting is therefore desirable. The authors previously proposed an energy harvester, called a flexible piezoelectric device (FPED), that consists of a piezoelectric film (polyvinylidene difluoride) and a soft material, such as silicon rubber or polyethylene terephthalate. The authors also proposed a system based on FPEDs for broadband energy harvesting. The system consisted of cantilevered FPEDs, with each FPED connected via a spring. Simply supported FPEDs also have potential for broadband energy harvesting, and here, a theoretical evaluation method is proposed for such a system. Experiments are conducted to validate the derived model.

Download Full-text